Skill: Audit Quality

What this Skill does

Perform comprehensive quality audit on knowledge base entries. Analyzes completeness, accuracy, documentation quality, and adherence to best practices.

Trigger

•User mentions "audit", "quality check", "review quality"
•Periodic KB maintenance (monthly/quarterly)
•Before major releases
•After bulk imports
•Part of curator workflow

What Claude can do with this Skill

1. Audit Single Entry

bash

# Detailed audit of one entry
python tools/kb.py validate python/errors/async-errors.yaml --audit

# Output shows:
# - Quality score breakdown
# - Missing fields
# - Improvement suggestions
# - Comparison to category average

2. Audit Category/Scope

bash

# Audit all Python entries
python tools/kb.py audit python/errors

# Audit all high-severity entries
python tools/kb.py audit --severity high

# Audit by scope
python tools/kb.py audit --scope universal

3. Comprehensive Audit

bash

# Audit entire KB
python tools/kb.py audit .

# Generates:
# - Overall quality report
# - Entries by score range
# - Categories needing attention
# - Improvement priorities

4. Quality Dimensions Audited

Dimension 1: Completeness (0-30 points)

code

✓ All required fields present (version, category, last_updated)
✓ Entry has id, title, severity, scope
✓ Problem statement clear and complete
✓ Solution has both code and explanation
✓ Prevention strategies included
✓ Tags are relevant and specific

Dimension 2: Technical Accuracy (0-30 points)

code

✓ Code examples are syntactically correct
✓ Solution actually solves the stated problem
✓ No deprecated APIs used
✓ Dependencies clearly stated
✓ Version compatibility noted
✓ Edge cases addressed

Dimension 3: Documentation (0-20 points)

code

✓ Problem description is clear
✓ Symptoms are reproducible
✓ Root cause is explained
✓ Solution explanation is thorough
✓ Real-world context provided
✓ Examples are practical

Dimension 4: Best Practices (0-20 points)

code

✓ Follows coding standards
✓ Security considerations included
✓ Performance implications noted
✓ Solution is maintainable
✓ Cross-references to related entries
✓ Prevention is actionable

5. Audit Report Format

code

📊 KB Quality Audit Report
Generated: 2026-01-07 14:45:32

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

OVERALL STATISTICS
───────────────────────────────────────────
Total Entries Audited: 127
Average Quality Score: 82.3/100
Score Range: 65 - 98

QUALITY DISTRIBUTION
───────────────────────────────────────────
Excellent (95-100):  12 entries (9.4%)   ⭐⭐⭐⭐⭐
Good (85-94):       45 entries (35.4%)  ⭐⭐⭐⭐
Acceptable (75-84): 48 entries (37.8%)  ⭐⭐⭐
Needs Work (65-74): 18 entries (14.2%)  ⭐⭐
Poor (<65):          4 entries (3.1%)    ⭐

BY SCOPE
───────────────────────────────────────────
universal:     85.2 avg (12 entries)  ⭐⭐⭐⭐
python:        83.1 avg (32 entries)  ⭐⭐⭐⭐
javascript:    80.5 avg (18 entries)  ⭐⭐⭐
docker:        78.9 avg (22 entries)  ⭐⭐⭐
postgresql:    84.3 avg (12 entries)  ⭐⭐⭐⭐
framework:     81.7 avg (28 entries)  ⭐⭐⭐⭐

TOP 10 ENTRIES
───────────────────────────────────────────
1. [UNIVERSAL-008] Testing Best Practices - 98/100 ⭐⭐⭐⭐⭐
2. [PYTHON-023] Async Context Managers - 96/100 ⭐⭐⭐⭐⭐
3. [DOCKER-015] Multi-stage Builds - 95/100 ⭐⭐⭐⭐⭐
... (7 more)

BOTTOM 10 ENTRIES (Need Attention)
───────────────────────────────────────────
1. [PYTHON-004] Import Errors - 65/100 ⭐
   Issues: Missing prevention, incomplete documentation
   Action: Enhance with examples and best practices

2. [DOCKER-007] Volume Mounts - 68/100 ⭐⭐
   Issues: Outdated examples, no edge cases
   Action: Update with current Docker practices

... (8 more)

IMPROVEMENT PRIORITIES
───────────────────────────────────────────
High Priority (Score < 75):
  - PYTHON-004: Import Errors (65) ⚠️
  - DOCKER-007: Volume Mounts (68) ⚠️
  - JAVASCRIPT-011: Promise Chains (70) ⚠️

Medium Priority (Score 75-84):
  - 48 entries need minor enhancements

Low Priority (Score 85-94):
  - 45 entries are good, could be excellent

RECOMMENDATIONS
───────────────────────────────────────────
1. Immediate: Improve 4 entries with score < 75
2. This Week: Enhance 18 entries in "Needs Work" range
3. This Month: Add examples to 48 "Acceptable" entries
4. Ongoing: Aim for 85+ average across all scopes

CATEGORY INSIGHTS
───────────────────────────────────────────
Strongest Categories:
  - Testing (avg: 87.3) ⭐⭐⭐⭐⭐
  - Architecture (avg: 86.1) ⭐⭐⭐⭐⭐

Weakest Categories:
  - Error Handling (avg: 76.8) ⭐⭐⭐
  - Performance (avg: 78.2) ⭐⭐⭐

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Key files to reference

•Quality rubric: @curator/QUALITY_STANDARDS.md
•Validation tool: @tools/validate-kb.py
•Entry format: @universal/patterns/shared-kb-yaml-format.yaml

Implementation rules

•Audit systematically - Don't skip entries
•Be thorough - Check all quality dimensions
•Provide feedback - Explain scores and suggestions
•Prioritize - Focus on low scores first
•Track progress - Re-audit after improvements

Common commands

bash

# Audit single entry (detailed)
python tools/kb.py validate entry.yaml --audit

# Audit category
python tools/kb.py audit python/errors

# Audit by severity
python tools/kb.py audit --severity high

# Audit entire KB
python tools/kb.py audit .

# Audit with verbose output
python tools/kb.py audit . --verbose

# Export audit report
python tools/kb.py audit . --output audit-report.md

Audit Workflow

Weekly Spot Checks

bash

# Audit 5 random entries from each scope
python tools/kb.py audit --random 5 --all-scopes

# Focus on recent additions
python tools/kb.py audit --since "2026-01-01"

Monthly Comprehensive Audit

bash

# Full KB audit
python tools/kb.py audit . --output monthly-audit-$(date +%Y%m).md

# Review and plan improvements
# Prioritize entries < 75 score

Quarterly Deep Dive

bash

# Complete audit + action plan
python tools/kb.py audit . --full-report

# Includes:
# - Quality trends over time
# - Category comparisons
# - Improvement tracking
# - Best practice identification

Quality Improvement Process

For Low-Scoring Entries (< 75)

•Identify gaps using audit report
•Research best practices (research-enhance skill)
•Add content to address gaps
•Re-validate to confirm improvement
•Track progress in metadata

For Mid-Range Entries (75-84)

•Polish documentation
•Add real-world examples
•Include edge cases
•Add performance notes
•Cross-reference related entries

For High-Scoring Entries (85+)

•Consider for promotion to universal scope
•Use as examples for new entries
•Identify patterns for reuse
•Share in team updates

Score Analysis

Score Trends

bash

# Track score changes over time
python tools/kb.py audit . --trend

# Compare to previous audit
python tools/kb.py audit . --compare previous-audit.md

Category Benchmarks

code

Excellent Categories (avg 85+):
  - Testing (87.3)
  - Architecture (86.1)
  - Security (85.8)

Target Categories (avg 80-84):
  - Async/Await (83.1)
  - Database (84.3)
  - CLI Tools (82.5)

Needs Improvement (avg < 80):
  - Error Handling (76.8) ← Focus here
  - Performance (78.2) ← Focus here
  - File I/O (79.1) ← Focus here

Related Skills

•kb-validate - Quick validation check
•find-duplicates - Check for duplicate content
•research-enhance - Enhance entries with research
•update-versions - Update outdated entries
•identify-gaps - Find missing content

Automation

Schedule automatic audits:

bash

# Weekly: Audit recent entries
0 9 * * 1 cd /path/to/kb && python tools/kb.py audit --since "1 week ago" --output weekly-audit.md

# Monthly: Full audit
0 9 1 * * cd /path/to/kb && python tools/kb.py audit . --output monthly-audit-$(date +\%Y\%m).md

Troubleshooting

Issue: "Audit takes too long"

Fix:

bash

# Audit smaller scope
python tools/kb.py audit python/errors

# Parallelize by scope
for scope in python javascript docker; do
  python tools/kb.py audit $scope &
done

Issue: "Scores seem wrong"

Check:

•Review quality rubric
•Verify validation rules
•Check for updated standards
•Re-run audit after changes

Issue: "Too many low scores"

Strategy:

•Prioritize by severity (critical > high > medium)
•Focus on frequently accessed entries
•Batch improvements by category
•Track and celebrate progress

Version: 1.0 Last Updated: 2026-01-07 Skill Type: Quality Assurance Target Score: 85/100 (average)